How the answers frame Danone
This snapshot, measured on 2026-09-29, covers 250 benchmark questions put to ChatGPT, Claude, Gemini, Perplexity and Grok, or 1,250 answers. None of the questions named a brand. Danone appeared in 59 of those answers. Grok named it most often (21 of 250 answers), and the other four assistants named it in 8 to 12 answers each.
The questions that brought Danone up were about packaging waste (26% visibility on that question intent), sourcing transparency (10%) and brands to avoid (9%). On reliability, value, safety and ingredient health, visibility was 2% or lower. The few answers tagged for what Danone was credited for point the same way: sustainability and ethics three times, customer experience once, and one answer with no substantive credit.
The five stored answers show how that plays out. Claude listed Danone among the major plastic polluters. Perplexity put it in a group accused of misleading recyclability claims, and noted that plastic-packaging lawsuits are allegations, not findings of lawbreaking. ChatGPT went the other way on a palm oil question. Its reasoning text said Danone and Nestlé disclose over 95% deforestation-free sourcing, and it listed Danone among companies making measurable progress. In Grok's deforestation answer Danone was among the brands named, but the visible excerpt highlighted Nestlé and Flora Food Group as leaders.
Perplexity also named Danone in a question about baby formula complaint handling. It cited a crisis-communications analysis of 2026 recalls that criticized Danone's messaging relative to Nestlé's.
…the 5 major plastic polluters are Coca-Cola, PepsiCo, Nestlé, Danone and Altria-Philip Morris International.
What the press is saying
PeakMetrics' news index logged 4,718 mentions of Danone in the last 30 days. Of those, 393 carried the brand in the headline.
The recent headlines are about product and corporate news. They include Danone publishing its 2025 UK nutrient profile score, a recall of a frozen dessert that could contain stones or other hard material, and the Steenvoorde site being positioned as a European medical nutrition hub with its largest production line in France. Another covers a plan with Intersnack to report on healthy food sales in the UK.
…accused by consumer groups of misleading claims about water bottles being “100% recycled” or recyclable.
The gap
The two stories barely overlap. The headlines sampled here are about nutrition reporting, manufacturing investment and a dessert recall. None of them is about packaging or deforestation, which are the topics that bring Danone up in the answers.
Volume does not explain it either. Among the 15 companies in this sector set, Danone's news count (4,718) is above Mars (1,859), Kimberly-Clark (2,334) and Hershey (2,097). It sits close to Mondelez (4,617) and Kraft Heinz (4,449). Yet Danone appeared in fewer answers (59) than each of those except Hershey (43), and far fewer than Procter & Gamble (314), Unilever (269) and Nestlé (266). Nestlé's press volume is about two and a half times Danone's. Its appearance count is more than four times higher. In this snapshot, press attention and AI visibility moved separately.
Where the answers do name Danone, they often name it beside the same peers (Nestlé, Coca-Cola, PepsiCo, Mondelez), in lists framed around plastic, greenwashing or deforestation.
…praised Nestlé’s consumer-facing updates and direct contact option, while criticizing Danone’s less consumer-focused messaging and lack of a clear support signpost.
What to watch
A comms team would take three things from this snapshot. First, the answers place Danone in environmental accountability conversations, and the same brand can appear as a plastic-pollution name in one answer and a deforestation-progress name in another. Second, Perplexity's recall comparison shows assistants drawing on third-party analysis of crisis communications. What gets said about a recall can reach answers about complaint handling. Third, the topics where Danone is nearly absent, such as ingredient health and safety, match the themes in its current press, which is a gap to test with more benchmark questions.
These are single runs of live answers, and results vary between runs. The five stored answers are a sample of the 59 appearances. Treat the figures as a dated baseline to re-measure, not a ranking or a trend.